We report here testing of CS with a rapid bio-assay system which has been used as a predictive screening test for carcinogenesis. This bio-assay system involves the application of the test compound to mouse skin, followed by the evaluation of the area of skin sebaceous gland non-specific esterase activity8"9 which can be accurately and rapidly obtained with an image analysing computer. The application of well-known carcinogens and tobacco condensates to mouse skin causes suppression of sebaceous gland non-specific esterase activity whereas the non-carcinogens and irritants, so far tested, do not possess this ability. The degree of suppression has also been found to correlate with the known potency of the carcinogenic compounds. CS is known to react readily with thiol groups at nearly neutral /?H (refs. 10-12). As this reaction could account for the suppression of non-specific esterase activity, the related a3-unsaturated compounds cinnamonitrile and cinnamaldehyde were also included in the experiments. The reactivity of these ap-unsaturated compounds with the thiol, glutathione, was in the order CS > cinnamaldehyde > cinnamonitrile.Groups of eight CFLP female mice, a hysterectomy-derived strain of Swiss origin 50 day old at start of treatment, when the hair growth cycle is in the telogen phase, were painted on three consecutive days with either cinnamonitrile, CS or cinnamaldehyde at 5 %, 1 % or 0.1 % concentrations in acetone. A control group of sixteen mice were treated with acetone alone. Three days after the final application all the mice were killed and a piece of skin from the centre of the painted area was excised and frozen in hexane pre-cooled to -75 C. Fig .1Behaviour in the non-specific esterase test of cinnamo-(o-o), cinnamaldehyde (o- o - o -o), and CS (o---o).The skin samples were mounted on cryostat tissue holders, and three 22 um transverse sections at 110 urn intervals were cut from each skin sample on a 'BrightV cryostat at a cabinet temperature of -25 C. The sections were air-dried at 6 C and fixed in formol calcium gum sucrose. Non-specific esterase activity was demonstrated in the sections by the method of Holt13 using 5-bromoindoxyl acetate as substrate. Sections of skin from every group were included in each incubation procedure to allow subsequent correction for incubation variation. The sections were then mounted in 90 % (w/v) aqueous solution of poly vinylpyrrolidone. The area of non-specific esterase activity in the sebaceous glands, as indicated by the reaction product, was measured in each skin sample using a 'Quantimet B' image analyser (Image Analysing Computers Ltd, Melbourne, Royston, Herts).The results (Fig. 1) show that of the three ap-unsaturated compounds examined only CS suppressed non-specific esterase activity. Because the relatively reactive cinnamaldehyde did not suppress the enzyme activity, reactivity with thiol groups alone does not explain the marked suppression caused by CS when applied to mouse skin at 1% and 5% (w/v) concentrations. In other comparative experiments the extent of suppression caused by CS was similar to that produced by 7,12-dimethylbenz[a]anthracene treatment.The reason why known carcinogens, or indeed CS, suppress non-specific esterase activity, at present, remains unknown. But considering the results of previous experiments using the sebaceous gland test and despite the obvious drawbacks of the predictive value of such bio-assay systems, CS should be fully investigated for carcinogenic activity.